312 CELESTIAL DYNAMICS. 



+ 6. Fisher, in lat. 80 N. and long. 11 E., noticed that 

 the surface-water had a temperature of 0, whilst at a depth 

 of 140 fathoms it stood at-f 8. 



As sea-water, unlike pure water, does not possess a point 

 of greatest density at some distance above the freezing-point, 

 and as the water in lat. 80 N. is found at some depth to be 

 warmer than water at the same depth 10 southward, we can 

 only explain this remarkable phenomenon of an increase of 

 temperature with an increase of depth by the existence of a 

 source of heat at the bottom of the sea. The heat, however, 

 which is required to warm the water at the bottom of an ex- 

 panse of ocean more than 1000 square miles in extent to a 

 sensible degree, must amount, according to the lowest esti- 

 mate, to some cubic miles of heat a day. 



The same phenomenon has been observed in other parts 

 of the world, such as the west coast of Australia, the Adri- 

 atic, the Lago Maggiore, &c. Especial mention should here 

 be made of an observation by Horner, according to whom the 

 lead, when hauled up from a depth varying from 80 to 100 

 fathoms in the mighty Gulf-stream off the coast of America, 

 used to be hotter than boiling water. 



The facts above mentioned, and some others which might 

 be added, clearly show that the loss of heat suffered by our 

 globe during the last 2500 years is far too great to have been 

 without sensible effect on the velocity of the earth's rotation. 

 The reason why, in spite of this accelerating cause, the length 

 of a day has nevertheless remained constant since the most 

 ancient times, must be attributed to an opposite retarding ac- 

 tion. This consists in the attraction of the sun and moon on 

 the liquid parts of the earth's surface, as explained in the 

 last chapter. 



According to the calculations of the last chapter, the re- 

 tarding pressure of the tides against the earth's rotation 

 would cause, during, the lapse of 2500 years, a sidereal day 

 to be lengthened to the extent of f 6 th of a second ; as the 



